Apify Easy Competitive Intelligence

Competitive Intelligence

Real-time competitive intelligence powered by live web data via Apify actors. Never answer competitive questions from training knowledge alone. Always gather live data first, then analyze.

Prerequisites

• Apify CLI v1.5.0+ (npm install -g apify-cli), or Apify MCP server
• Authenticated session (apify login or APIFY_TOKEN env var)

CLI rules: Always pass --json, --user-agent apify-awesome-skills/apify-easy-competitive-intelligence, and 2>/dev/null.

• Run actor: apify actors call "ACTOR_ID" -i 'INPUT' --json 2>/dev/null → returns run metadata with defaultDatasetId
• Fetch results: apify datasets get-items DATASET_ID --format json > /tmp/results.json 2>/dev/null — save locally, parse from file:
• Quick extraction: jq '.[] | "\(.field1) | \(.field2)"' /tmp/results.json
• Aggregation: python3 -c "import json; d=json.load(open('/tmp/results.json')); ..."
• Tabular: --format csv > /tmp/results.csv + python3 with csv.DictReader
• Flags: --limit N, --offset N, --format json|jsonl|csv|xlsx|xml
• Output fields: apify datasets info DATASET_ID --json | jq .fields
• Fetch schema: apify actors info "ACTOR_ID" --input --json 2>/dev/null

If CLI is unavailable and Apify MCP server is connected, use MCP call-actor / fetch-actor-details / get-actor-output directly.

Authentication

If a CLI command fails with an auth error, authenticate using one of these methods:

1. OAuth (interactive): apify login (opens browser) 2. Environment variable: export APIFY_TOKEN=your_token_here 3. From .env file: source .env (if the file contains APIFY_TOKEN=...)

Generate token: https://console.apify.com/settings/integrations

Actor Registry

Every actor call follows three steps: 1. Read — find the actor's section in reference/actor-schemas.md. Use the exact verified input and follow the "How to find" instructions for URLs/slugs. 2. Discover — verify platform URLs and slugs (e.g. via SERP) as described in the actor's schema section. Do not guess — wrong slugs silently return empty or wrong data. 3. Run — call the actor with verified input.

Alternatively, fetch the live schema: apify actors info "ACTOR_ID" --user-agent apify-awesome-skills/apify-easy-competitive-intelligence --input --json 2>/dev/null

Core Workflow

Step 0: Understand the User (once, at start)

Clarify before gathering data:

• Role — Analyzed company, competitor, investor, consultant?
• Decision — Entering market, defending position, choosing vendor, building battlecard?
• Autonomy — Checkpoints after initial findings, or autopilot?

Steps 1–7

1. Clarify scope — Identify competitors. Select module(s). Default geography: US. 2. Read module reference — Load reference/modules/<module>.md for gathering + analysis instructions. 3. Gather live data — For each actor call, follow the three-step pattern: Read (actor-schemas.md) → Discover (SERP for URLs) → Run (call actor). Use PRIMARILY actors from the Actor Registry above. 4. Checkpoint (if not autopilot) — Present first findings, confirm direction. 5. Analyze — Select framework, lead with narrative, support with tables. 6. Verify — Run pre-delivery verification (reference/verification-checklist.md). Check: every claim has a source URL, every major finding has a confidence label, inferences are labeled as such. Remove any ungrounded claims. 7. Deliver — End with strategic recommendations framed for the user's role.

Framework Selection

Data Collection Rules

• Prefer structured actors over website-content-crawler when a dedicated actor exists.
• Cost budget — 3-8 actor calls per snapshot. Track total, warn at 15+.
• Parallelize independent call-actor calls in a single response.
• Failures — Report every failure explicitly (actor, input, error). Retry with corrected input if the cause is obvious. If retry fails, try rag-web-browser as fallback. Never silently skip a failed data source.
• Cite everything — Include source URLs for every data point.
• Async for long runs — Set async: true for actors >30s, poll with get-actor-run.
• Protected platforms — Do NOT use website-content-crawler or rag-web-browser for: g2.com, capterra.com, gartner.com, glassdoor.com, reddit.com, linkedin.com. Use dedicated actors.

Apify vs. WebSearch

Apify required: review sites (G2, Capterra, Gartner, Glassdoor), LinkedIn, Reddit, Amazon, Walmart, app stores, SimilarWeb, Crunchbase, Wayback Machine, Google Maps reviews, news (Google News actor).

WebSearch/WebFetch sufficient (Claude Code built-in tools): competitor discovery, general company info, blog posts, publicly accessible pricing pages.

Data Validation & Grounding

• Every factual claim needs a source URL. No link = not a fact.
• Confidence labels are mandatory. Mark every major finding: High (primary source), Medium (2+ third-party sources), Low (single third-party source). Format: [Confidence | Source]. No report without labels.
• Data tiers: Verified (primary source) → Reported (third-party, attribute) → Inferred (label as "this suggests...") → Ungrounded (omit).
• Numbers are dangerous — employee counts, revenue, funding change fast. Always cite source and date.
• Empty results ARE intelligence — 0 jobs = not hiring, 0 SimilarWeb = small site, 12 reviews = low adoption.
• Cross-reference — Single-source claims are unverified. Multi-source (G2 + Capterra + Reddit) = pattern.

Module Selection

Actor Input & Output Schemas

Verified inputs and output fields from real test runs. Last tested: 2026-05-06.

---

apify/google-search-scraper

Input:

{ "queries": "\"Company Name\" keyword", "maxPagesPerQuery": 1 }

Optional: countryCode ("us"), languageCode ("en"), maxPagesPerQuery (1-10).

Output keys: searchQuery, url, organicResults[], paidResults[], peopleAlsoAsk[], relatedQueries[], aiOverview, resultsTotal

Each organicResult: title, url, description, position, date, emphasizedKeywords[], siteLinks[], productInfo (may contain rating, numberOfReviews).

---

apify/website-content-crawler

Input:

{
  "startUrls": [{"url": "https://example.com/pricing"}],
  "maxCrawlPages": 1,
  "maxCrawlDepth": 0,
  "proxyConfiguration": {"useApifyProxy": true}
}

⚠️ proxyConfiguration is required. Optional: crawlerType ("playwright:adaptive").

Output keys: url, markdown, text, html, metadata, crawl, screenshotUrl

metadata: title, description, author, keywords, languageCode, openGraph, jsonLd. crawl: loadedUrl, loadedTime, httpStatusCode, depth, contentType.

---

apify/rag-web-browser

Input:

{ "query": "Company Name pricing plans" }

Output keys: query, markdown, metadata, searchResult, crawl

searchResult: title, description, url, resultType, rank. metadata: title, description, languageCode, url. markdown: full page content in markdown.

---

dev_fusion/Linkedin-Company-Scraper

⚠️ Before calling: LinkedIn slugs often differ from company names (e.g., Oxylabs → oxylabs-io, not oxylabs). Wrong slug silently returns 0 results. Discover via SERP first: "[company] site:linkedin.com/company"

Input:

{ "profileUrls": ["https://www.linkedin.com/company/VERIFIED-SLUG/"] }

⚠️ Field is profileUrls, NOT urls.

Output: Data stored in key-value store, not dataset. Check KV store keys after run.

---

curious_coder/linkedin-jobs-scraper

⚠️ Before calling: Verify the company's exact name on LinkedIn via SERP: "[company] site:linkedin.com/jobs". Multi-word names need URL-encoding (Bright Data → Bright%20Data). Wrong name returns 0 or unrelated jobs silently.

Input:

{
  "urls": ["https://www.linkedin.com/jobs/search/?keywords=VERIFIED-COMPANY-NAME&position=1&pageNum=0"],
  "count": 10,
  "scrapeCompany": true
}

⚠️ count minimum is 10. Requires LinkedIn search URL, NOT keyword string.

Output keys: id, title, companyName, companyLinkedinUrl, companyLogo, companyWebsite, companyDescription, companyEmployeesCount, companySlogan, location, country, postedAt, postedAtTimestamp, expireAt, salary, salaryInsights, seniorityLevel, employmentType, jobFunction, industries, descriptionText, descriptionHtml, applicantsCount, applyUrl, applyMethod, workplaceTypes, workRemoteAllowed, standardizedTitle, link, inputUrl

---

pratikdani/crunchbase-companies-scraper

⚠️ Before calling: Verify the org slug via SERP: "[company] site:crunchbase.com/organization". Slug is typically lowercased with hyphens.

Input:

{ "url": "https://www.crunchbase.com/organization/VERIFIED-SLUG" }

⚠️ Field is url (singular string), NOT urls (array).

Output: May return {"error": "Issue in running the url."} for wrong slugs. When successful: company profile data (funding, investors, financials).

---

junglee/Amazon-crawler

Input:

{ "categoryOrProductUrls": [{"url": "https://www.amazon.com/dp/ASIN"}] }

⚠️ Field is categoryOrProductUrls, NOT productUrls.

How to find the URL: Any Amazon product page works. The canonical form is amazon.com/dp/ASIN where ASIN is the 10-character product identifier (e.g., B0D1XD1ZV3). ASIN is visible in the URL or in the product details section of any Amazon listing.

Output keys: title, url, asin, price, listPrice, brand, stars, reviewsCount, inStock, features[], attributes[], productOverview[], description, thumbnailImage, highResolutionImages[], bestsellerRanks, seller, delivery, aiReviewsSummary, productPageReviews[], monthlyPurchaseVolume, variantAsins[]

---

web_wanderer/amazon-reviews-extractor

Input:

{ "products": ["https://www.amazon.com/dp/ASIN"] }

⚠️ Field is products, NOT productUrls. Use the same amazon.com/dp/ASIN format as Amazon-crawler.

May return 0 items for some products — try the full product URL with title slug if the short form fails.

---

e-commerce/walmart-product-detail-scraper

Input:

{ "productUrls": ["https://www.walmart.com/ip/Product-Name/ID"] }

How to find the URL: Use the full URL from Walmart product page. The numeric ID at the end is required (e.g., /1752657021).

⚠️ May return 0 items. Verify product URL is accessible (some products are region-locked).

---

compass/Google-Maps-Reviews-Scraper

Input:

{ "startUrls": [{"url": "https://www.google.com/maps/place/Place+Name/@lat,lng,17z"}], "maxReviews": 30 }

How to find the URL: Search Google Maps for the business → copy full URL from browser address bar. Must include the @lat,lng,zoom part. Alternatively use a place ID URL: https://www.google.com/maps/place/?q=place_id:ChIJ....

Output: Status object with isFinished, enqueued, placeIdsEnqueued. Review data in separate dataset items.

---

automation-lab/g2-scraper

Replaces zhorex/g2-reviews-scraper (broken).

⚠️ Before calling: Discover the G2 slug via SERP: "[product] site:g2.com/products". Extract slug from URL pattern g2.com/products/[slug]/reviews. Wrong slug silently returns reviews for a different product (e.g., Slack). Always verify productName in output matches your target.

Input:

{ "mode": "product_reviews", "productUrls": ["VERIFIED-SLUG"], "maxReviews": 25, "sortReviews": "newest" }

⚠️ Field is productUrls (array of slugs), NOT startUrls. mode is required.

Output keys: reviewId, title, starRating, nps, reviewText, publishedAt, submittedAt, reviewerName, country, region, easeOfUse, easeOfSetup, easeOfAdmin, qualityOfSupport, meetsRequirements, loveTheme, hateTheme, switchedFromOtherProduct, switchedReason, companySegment, industry, productName, productSlug, url, helpfulVotes, sourceType

---

zen-studio/capterra-reviews-scraper

⚠️ Before calling: Discover the Capterra URL via SERP: "[product] site:capterra.com/p/". URL pattern is capterra.com/p/[numeric-id]/[Product-Name]/reviews/. The numeric ID is required — do not guess it.

Input:

{ "productUrl": "https://www.capterra.com/p/NUMERIC-ID/Product-Name/reviews/", "maxReviews": 10 }

⚠️ Field is productUrl (singular string), NOT startUrls.

Output keys: url, reviewId, title, writtenOn, overallRating, easeOfUseRating, customerSupportRating, functionalityRating, valueForMoneyRating, recommendationRating, prosText, consText, generalComments, adviceToOthers, incentivized, reviewer, vendorResponse, scrapedAt

reviewer: fullName, jobTitle, companySize, industry, timeUsedProduct, isValidated.

---

Gartner Peer Insights

⚠️ No working actor available (tested 2026-05-06 with Databricks — major vendor with thousands of reviews). All actors on Apify Store (zen-studio, hello.datawizards, memo23) return empty or fail. Gartner blocks all scrapers. Use SERP snippet mining as fallback:

call-actor: apify/google-search-scraper
  input: { "queries": "[product] review site:gartner.com" }

---

memo23/glassdoor-scraper-ppr

⚠️ Before calling: Discover the Glassdoor URL via SERP: "[company] site:glassdoor.com/Overview". The URL contains an employer ID (EI_IE[number]) — the numeric ID is what matters, not the company name part.

Input:

{ "startUrls": [{"url": "https://www.glassdoor.com/Overview/Working-at-COMPANY-EI_IEVERIFIED-ID.htm"}], "command": "reviews" }

command values: reviews, jobs, interviews, salaries, overview. Default to reviews for CI.

Output keys: reviewId, summary, pros, cons, advice, ratingOverall, ratingWorkLifeBalance, ratingCultureAndValues, ratingCompensationAndBenefits, ratingCareerOpportunities, ratingSeniorLeadership, ratingDiversityAndInclusion, ratingCeo, ratingBusinessOutlook, ratingRecommendToFriend, jobTitle, location, reviewDateTime, isCurrentJob, lengthOfEmployment, employer

employer: id, shortName, squareLogoUrl. jobTitle: text.

---

harshmaur/reddit-scraper

Input:

{ "startUrls": [{"url": "https://www.reddit.com/search/?q=company+keyword"}], "maxItems": 10 }

Alternatively, scrape a specific subreddit: https://www.reddit.com/r/subreddit/search/?q=keyword.

Output keys: id, title, body, authorName, communityName, upVotes, commentsCount, dataType ("post"/"comment"), postUrl, contentUrl, createdAt, crawledAt, flair, postType

---

neatrat/google-play-store-reviews-scraper

Input:

{ "appIdOrUrl": "com.company.app" }

⚠️ Field is appIdOrUrl, NOT appId.

How to find the app ID: Open the app on Google Play → the URL is play.google.com/store/apps/details?id=com.company.app. The id parameter is the app ID (e.g., com.slack, com.spotify.music). Both the ID string and the full Play Store URL work as input.

---

jdtpnjtp/apple-app-store-scraper

⚠️ Requires SHADER proxy group which may not be available on all Apify plans. Verify access before use.

Input (when proxy available): App Store URL, e.g., https://apps.apple.com/us/app/app-name/idNUMBER.

How to find the URL: Search App Store or use SERP: "[app] site:apps.apple.com". The numeric ID at the end (id618783545) identifies the app.

---

pro100chok/similarweb-scraper

Input:

{ "searchType": "similarweb", "domains": ["apify.com", "oxylabs.io", "brightdata.com", "zyte.com", "scraperapi.com", "smartproxy.com", "scrapingbee.com", "nimbleway.com", "diffbot.com", "octoparse.com"] }

⚠️ searchType is required. Minimum 10 domains required — actor rejects fewer with "Minimum 10 domains required for analysis." Always batch all competitors into one call. Use bare domain without protocol (e.g., apify.com, not https://apify.com).

---

data_xplorer/google-news-scraper-fast

Input:

{
  "keywords": ["\"Company Name\""],
  "maxArticles": 10,
  "timeframe": "7d",
  "region_language": "US:en",
  "decodeUrls": true,
  "extractDescriptions": true,
  "extractImages": false
}

timeframe values: 1h, 1d, 7d, 1y, all. Multi-keyword via array. Use quotes for exact phrase match (e.g., "\"Bright Data\"" to find articles mentioning "Bright Data" as a phrase, not "bright" and "data" separately). ⚠️ Boolean operators (OR, AND) produce unreliable results — use separate keywords array entries instead of boolean syntax.

Output keys: title, url, source, publishedAt (ISO), publishedTimestamp (unix), image, description (full text when extractDescriptions: true), metadata

metadata: scrapeTimestamp, keyword, sourceType, timeframe.

---

andok/wayback-machine-scraper

Input:

{ "url": "https://example.com/pricing" }

No maxSnapshots parameter exists. Use full URL including path to specific page (e.g., pricing page, homepage).

Output keys: inputUrl, snapshotCount, snapshots[], latestSnapshot, latestHtml, checkedAt, error

Each snapshot contains URL + timestamp of archived version. May return "error": "Wayback CDX HTTP 503" when the Wayback Machine API is overloaded — retry later.

Competitor Snapshot

When to use: Analyze, profile, or understand a specific competitor.

Data Gathering

# 1: Discover website + news
call-actor: apify/google-search-scraper  # "[competitor name]" branded query

# 2: Scrape key pages (parallel)
call-actor: apify/website-content-crawler  # homepage, /pricing, /about

# 3: Structured enrichment (parallel, if URLs available)
call-actor: dev_fusion/Linkedin-Company-Scraper
call-actor: pratikdani/crunchbase-companies-scraper

# 4: Temporal context — what changed?
call-actor: andok/wayback-machine-scraper  # homepage + pricing ~1 year ago

# 5: Recent news — last 7 days
call-actor: data_xplorer/google-news-scraper-fast

Analysis

Synthesize into: positioning, target audience, key claims, strengths, vulnerabilities. Compare to user's product if context available. Highlight changes over time (Wayback) and recent momentum (news).

Content & SEO Battle

When to use: Understand competitors' content strategy or search positioning.

Data Gathering

# 1: SERP rankings for category keywords (parallel, 4-6 buyer keywords, NOT brand terms)
call-actor: apify/google-search-scraper
  input: { "queries": "[category] tools comparison [current-year]\nbest [category] platform\n[category] alternatives\n[use case] software", "maxPagesPerQuery": 1 }

# 2: Indexed content volume
call-actor: apify/google-search-scraper
  input: { "queries": "site:[competitor.com]\nsite:[competitor-2.com]" }

# 3: Traffic data (may be empty for small sites)
call-actor: pro100chok/similarweb-scraper  # minimum 10 domains — batch all competitors in one call

# 4: Blog/content strategy
call-actor: apify/website-content-crawler  # [competitor-url]/blog, crawl 10 pages

The most valuable SEO finding is often absence — not ranking for category keywords is a critical gap.

Analysis

Map competitors to keywords. Estimate content volume and frequency. Identify topic clusters. Find content gaps — topics nobody covers well that user could own.

Hiring Signal Analysis

When to use: Infer competitor's strategic direction from hiring patterns.

Data Gathering

# 1: LinkedIn job listings
call-actor: curious_coder/linkedin-jobs-scraper

# 2: Fallback — careers page
call-actor: apify/website-content-crawler  # [competitor-url]/careers

# 3: Glassdoor — culture, salaries, internal signals
call-actor: memo23/glassdoor-scraper-ppr

# 4: Recent hiring news
call-actor: apify/google-search-scraper
  input: { "queries": "[competitor] hiring jobs careers [current-year]\n[competitor] layoffs OR expansion [previous-year] [current-year]" }

0 LinkedIn results = signal (not hiring aggressively). Glassdoor compensates — reviews reveal culture/strategy even without active hiring.

Analysis

Categorize roles by department. Hiring velocity (scaling/stable/contracting). Technology signals from JDs. Geographic expansion. Seniority mix: hiring leaders = new initiative, hiring ICs = scaling existing.

Market Landscape Map

When to use: Understand all players, find white space, or map the competitive landscape.

Data Gathering

# 1: Discover players (parallel)
call-actor: apify/google-search-scraper
  input: { "queries": "[industry] companies\nbest [category] tools\n[category] alternatives" }

# 2: Category/comparison pages
call-actor: apify/rag-web-browser
  input: { "query": "best [category] comparison" }

# 3: Quick snapshot of each discovered competitor
call-actor: apify/website-content-crawler  # homepage of each discovered competitor

# 4: Funding/size enrichment
call-actor: pratikdani/crunchbase-companies-scraper

# 5: Traffic comparison
call-actor: pro100chok/similarweb-scraper  # minimum 10 domains — batch all competitors in one call

Analysis

Categorize by tier (enterprise, mid-market, SMB, open-source). Build positioning map (e.g., price vs. feature breadth). Identify white space — underserved segments or unowned positioning. Note trends, recent entrants, consolidation signals.

Pricing Deep Dive

When basic Pricing Intelligence isn't enough — complex pricing models, multiple products/SKUs, or per-use-case/per-tier comparison needed.

When to Use

• Basic pricing scrape revealed non-trivial structure (usage-based, credit multipliers, tiered)
• User wants deeper comparison than headline price
• Multiple products/SKUs from one vendor with different pricing

Pricing Decomposition (Nagle)

For each vendor, decompose pricing into 4 components (ref: Thomas Nagle, The Strategy and Tactics of Pricing):

1. Price metric — unit of purchase (seat, API call, record, GB, credit, transaction) 2. Price structure — how price varies (flat, tiered, volume, usage-based, credit × multiplier, commitment terms) 3. Offering design — how features are bundled (all-inclusive, modular, add-on, freemium, good/better/best) 4. Price fences — what separates segments (feature gates, support tiers, SLA, SSO, compliance, API access)

This decomposition reveals hidden costs and comparison dimensions.

Normalization

Based on the decomposition:

1. Identify common price metric — what unit makes sense for this category? 2. Propose comparison dimensions — which axes matter? (per tier, per use case, per product line) 3. ⚠️ Confirm with user — always, even in autopilot mode, unless normalization was specified by the user in the prompt. Explain and confirm the decomposition principles and proposed comparison dimensions to the user. 4. Normalize — convert each vendor to common unit 5. Flag incomparables — where and why the comparison breaks down

Confidence

• V (Verified) — directly from pricing page / docs
• C (Calculated) — derived from documented values (base × multiplier)
• E (Estimated) — indirect estimate from plan names, third parties
• ? (Contact sales) — vendor doesn't publish

Pricing Intelligence

When to use: Compare pricing, understand pricing models, or find positioning opportunities.

Data Gathering

# Scrape pricing pages (parallel)
call-actor: apify/website-content-crawler  # [competitor]/pricing for each vendor

# E-commerce products (if applicable)
call-actor: junglee/Amazon-crawler
call-actor: e-commerce/walmart-product-detail-scraper

# Third-party pricing breakdowns
call-actor: apify/google-search-scraper
  input: { "queries": "[competitor] pricing review" }

Analysis

Extract plan names, prices, feature lists, limits. Normalize into comparison matrix. Identify pricing model types (per-seat, usage-based, freemium, enterprise-only). Flag positioning signals — gaps, undercut potential, differentiation opportunities.

Review Intelligence

When to use: Understand customer sentiment, competitor pain points, or exploitable gaps.

SERP Snippet Mining (Always Start Here)

SERP results for review sites contain averageRating, numberOfReviews, and pros/cons snippets — fast triage before full scraping.

call-actor: apify/google-search-scraper
  input: { "queries": "[company] review pros cons site:g2.com\n[company] review site:capterra.com\n[company] vs [rival] review", "maxPagesPerQuery": 1 }

Full Review Scraping

Use dedicated actors only — review sites block generic scrapers.

# Structured reviews (if applicable)
call-actor: automation-lab/g2-scraper
call-actor: zen-studio/capterra-reviews-scraper
# Gartner: no working actor — use SERP snippet mining above as fallback

# Unfiltered community sentiment (if applicable)
call-actor: harshmaur/reddit-scraper

# Product/location reviews (if applicable)
call-actor: compass/Google-Maps-Reviews-Scraper
call-actor: web_wanderer/amazon-reviews-extractor
call-actor: neatrat/google-play-store-reviews-scraper
call-actor: jdtpnjtp/apple-app-store-scraper

Analysis

Categorize sentiment. Extract top praised features, top complaints, feature requests. Identify switching signals ("switched from X", "better than Y"). Complaints = positioning opportunity.

Multi-Module Playbook

For comprehensive analysis (battlecard, deep dive, board prep). Uses subagent architecture — each module runs as independent agent.

Module Sequence

Each step builds on prior findings (read definitions in modules/):

1. Market Landscape → competitor list + URLs + positioning signals 2. Competitor Snapshots → profiles using URLs from step 1 3. Pricing Intelligence → gathered alongside snapshots (pricing pages) 4. Review Intelligence → uses company names from step 1 for SERP queries 5. Hiring Signals → uses company names + career URLs from steps 1-2 6. Content & SEO → uses category keywords + domains from step 1

Data Carry-Forward

• URLs from Market Landscape → reuse everywhere (avoid re-searching)
• Company names → Review SERP queries + Hiring LinkedIn URLs
• Positioning insights from Snapshots → inform SEO keywords
• Review complaints → validate/contradict Snapshot claims

Subagent Architecture

Run discovery SERP directly (small, fast), then spawn parallel subagents:

• Competitor Profiler (Snapshots + Pricing) — URLs, pricing pages
• Review Analyst — company names, main rivals
• Hiring & Culture Analyst — company names, career URLs
• SEO & Content Analyst — domains, category keywords

Each subagent brief must include: user context (Step 0), competitor list + URLs, module reference path, output format (summary under 500 words: headline finding, key data table, 3 insights with source URLs).

Synthesis

1. Executive narrative — 1 paragraph connecting biggest cross-module finding 2. Top 5 cross-module insights (e.g., "G2 reviews confirm pricing vulnerability from snapshot") 3. Competitive comparison matrix 4. Strategic recommendations — prioritized, framed for user's role 5. Data gaps and confidence levels

Pre-delivery Verification Checklist

Run this audit on every claim before presenting the final report.

Audit Steps

1. Source audit — Every factual statement has a source URL? If not, find it in scraped data or remove the claim. 2. Data-back check — Every insight traces to specific data points. "Weak SEO" must be backed by "absent from top 10 for [keyword] — [URL]". Insight without data = opinion — trace back or remove. 3. Contradiction scan — Claims across modules contradict? (e.g., snapshot says "strong support" but G2 shows complaints). Flag and explain, don't hide.

Confidence Labeling

Mark each major finding:

Label format: [Confidence | Source] — e.g., "500+ connectors [High | Primary — fivetran.com/connectors]"

No inferences in reports. Untraceable claims must be removed.